It is estimated that bearing faults account for around half of all failures in electric machines. The consequences of failure can result in unplanned downtime, machine breakdowns, equipment damage, and financial losses. To prevent this and ensure consistent performance, effective bearing maintenance is essential. But, when a single machine can contain hundreds of bearings, how can this be managed? Here, Chris Johnson, managing director of bearing supplier SMB Bearings explains.
Of the often hundreds of bearings found in machines, most of them will outlive the machine itself. In fact, only a very small fraction of bearings actually fail in service. Unfortunately, when they do, the impact can be significant. Having an effective condition monitoring strategy in place can help to mitigate this risk.
Vibration monitoring and analysis is one of the most common ways to measure bearing performance and to implement predictive maintenance. Machinery emits vibration signals, as does the movement and interaction between components within the bearing itself. Any changes in these signals that aren’t caused by changes to operational conditions can indicate an issue with machine health.
Typically faults in moving components such as bearings are responsible for signal changes, and these changes can be detected using equipment such as accelerometers and vibration velocity transducers. These defective signals may be caused by geometric imperfections, surface roughness, improper lubrication, or contamination.
Vibration monitoring is an acceptable and reliable form of condition monitoring, with standards such as ISO 10816 available for vibration evaluation. When implemented correctly it can be an extremely effective way to detect bearing faults, but it can be costly to set up and requires direct access to the machine to install sensors.
It may also not be suitable for low-speed machines or environments with high levels of noise, where defective signals may be weaker and harder to detect. This has been found to be the case in applications such as turbines, engines, and transmissions. For these cases, other methods might be more appropriate.
In environments with high levels of noise, low-frequency vibrational changes caused by such small parts are often almost undetectable compared to the background noise and signals. However, these defects produce waves in higher frequency ranges(100kHz) known as acoustic emission (AE).
These transient elastic waves are produced by the release of energy caused by defects on the surface of a material or component. They can be generated by faulty or damaged bearings and picked up by an AE transducer with little interference from surrounding machinery. Though this can be a little more costly and the signals require expert analysis, this method can provide a high signal-to-noise ratio and be used in a range of environments to monitor bearing condition.
Bearings often rely on oil and grease to lubricate them and ensure their function. Exposure to heat can degrade lubricants and produce chemical by-products. General wear can also result in the production of debris. Oils and greases can be analyzed to examine the debris present and to determine whether it has resulted from the wear or the bearing itself, which may suggest a wider fault and can be used to monitor condition.
Performing chemical analysis on oils and greases is a simple and effective way to detect degradation, but this method is most suited to larger machines with lubricated bearings.
The Institute of Electrical and Electronics Engineers standard IEEE 841 states that at a rated load, the stabilized bearing temperature rise should be no more than 45°C. Many factors can influence temperature rise, including bearing or lubricant degradation, operational speeds, or the temperature within the motor itself.
Monitoring this temperature increase for unusual levels can alert engineers to a fault within the bearing and allow for investigation. This method is traditional and effective as it allows quantitative measurement of specific parameters, but it does require the installation of embedded temperature sensors.
Stator current monitoring
Current monitoring is most often used for measuring the condition of bearings in electric machines and motors. A faulty bearing in a motor can cause changes in flux density, which will lead to stator current variation. This variation can be measured to identify bearing faults and is very straightforward and non-invasive to implement.
The required parameters can be measured using existing systems in electromechanical operations, meaning that additional sensors are often not necessary, and the implementation cost is lower. In some cases, it can be carried out remotely from a control room.
Measuring stator current can also give a wider overview of machine health, rather than just that of the bearings. The main limitation is that signal-to-noise ratio can be low, and signatures are sometimes harder to detect in some applications. Though research has identified frequencies associated with single-point faults, more general issues such as roughness are not as well characterized, but overall this is an effective method for electric machinery and motor systems.
Measuring infrared emission can give insight into bearing performance. Using thermography equipment or infrared cameras can measure the release of infrared energy and highlight areas where temperatures vary. Improper lubrication and other faults impacting speed or load can induce heat variation or overheating and using infrared can visually identify areas for concern, allowing engineers to see the exact location of the potential fault.
This technique does rely on the defect producing excess heat, meaning that it is not always sensitive enough to detect early issues or those not severe enough to generate infrared emissions, but it does allow for rapid analysis of bearing condition in many cases.
With bearings such a crucial component of many machines, monitoring their condition effectively should be a priority for any industrial operation. There are many techniques available to achieve this depending on the industrial operation, the type of machinery used, and the equipment available.
The best method to monitor bearing condition will depend on the operation itself. Implementing a suitable method to measure bearing health and to detect and identify faults before they become more serious is an invaluable way to ensure bearing performance and to prevent failures, costly breakdowns, and unplanned downtime.